CN113676926B - User network perception profiling method and device - Google Patents

Abstract

The embodiment of the invention provides a user network perceived image method and a device, wherein the method comprises the following steps: collecting internet surfing behavior data, XDR data, MR data and complaint data of a target user, and analyzing performance data representing network perception of the target user from the internet surfing behavior data; correlating the performance data, XDR data, MR data and complaint data, obtaining perception fusion information of the target user, and obtaining the service type of the perception fusion information according to the perception fusion information; selecting data corresponding to the service type from the perception fusion information according to the service type; inputting the selected data into a user perception portrait model, outputting the network perception score of the target user, and obtaining the user network perception description corresponding to the network perception score. The embodiment of the invention realizes accurate user perception portrait.

Description

User network perception profiling method and device

Technical field

The invention belongs to the technical field of mobile communication wireless networks, and in particular relates to a user network sensing portrait method and device.

Background technique

Existing technologies mainly rely on user surveys, user complaints and monitoring network indicators to simulate and evaluate user network perception and locate poor network quality problems.

Among them, user survey methods generally use scores or simple descriptions to evaluate perceptions due to the strong subjectivity of users. The dimensions are vague, single and poorly professional, making it difficult to obtain accurate user perception evaluations and locate the root causes of poor network quality.

The user complaint method is lagging. Only after the user complains can we know the user perception and the location of poor network quality problems, which seriously affects the user experience. In addition, users' descriptions of problem phenomena and locations are often vague, making it difficult to accurately locate the cause of poor quality and the involved network elements or cells.

Monitoring network indicator method Since network indicators are statistical averages at the network element level, and because the minimum granularity of network indicator statistics is minute level, and the granularity of user perception anomalies are second-level events, it is difficult to monitor network indicators to accurately map the perception of a single user. , currently there is a common situation where the indicators are excellent but the user perception is not good. In addition, the flexibility and diversity of Internet applications and the lack of effective mapping between user experience and telecommunications protocols are also important reasons why it is difficult for measurement indicators to match user perceptions.

Contents of the invention

In order to overcome the above-mentioned problems that the existing user network-aware profiling method produces inaccurate and hysteretic user network-aware portraits or at least partially solve the above problems, embodiments of the present invention provide a user network-aware profiling method and device.

According to a first aspect of an embodiment of the present invention, a user network awareness profiling method is provided, including:

Collect the target user's online behavior data, XDR data, MR data and complaint data, and parse the performance data that characterizes the target user's network perception from the online behavior data;

Correlate the performance data, XDR data, MR data and complaint data to obtain the perceptual fusion information of the target user, and obtain the service type of the perceptual fusion information according to the perceptual fusion information;

Select the data corresponding to the service type from the perception fusion information according to the service type; input the selected data into the user perception portrait model, output the network perception score of the target user, and obtain the network perception score corresponding to the network perception score. User network awareness description;

Wherein, the service type and the data corresponding to the service type are stored in association in advance;

The network awareness score and the user network awareness description are stored in association in advance;

The user perception portrait model is trained and acquired based on the perception fusion information of the user sample and the network perception score of the user sample.

Specifically, the steps for collecting target users’ online behavior data include:

Collect log data of the target users through SDK burying method;

Crawl the network data of the target user through a web crawler or public API;

The collected log data and captured network data are used as the target user's online behavior data.

Specifically, the steps of correlating the performance data, XDR data, MR data and complaint data to obtain the perceptual fusion information of the target user include:

According to the timestamp, IMEI, IMSI, MME UES1AP ID, Cell ID and eNB ID fields that exist in the XDR data, MR data and complaint data, associate the XDR data, MR data and complaint data to obtain the Related fusion data of target users;

According to the timestamp, App Type, App Sub-type and IMEI fields that exist in the performance data and the associated fusion data, the performance data and the associated fusion data are associated to obtain the perceptual fusion information of the target user.

Specifically, the step of inputting the selected data into the user perception portrait model and outputting the network perception score of the target user also includes:

The business type of obtaining the perceptual fusion information of each user sample based on the perceptual fusion information of each user sample;

According to the service type of the perceptual fusion information of all the user samples, divide the perceptual fusion information of all the user samples into multiple data subsets;

For any of the data subsets, select the data corresponding to the business type from the data subset according to the business type to which the data subset belongs;

Train the user perception profile model based on the data selected from the data subset, and obtain the user perception profile model corresponding to the business type to which the data subset belongs;

Accordingly, the step of inputting the selected data into the user perception portrait model and outputting the network perception score of the target user includes:

According to the service type of the target user's perception fusion information, obtain the user perception portrait model corresponding to the service type;

The data selected from the perceptual fusion information of the target user is input into the user perceptual portrait model corresponding to the service type, and the network perceptual score of the target user is output.

Specifically, the user perception profile model is trained based on the data selected from the data subset, and the step of obtaining the user perception profile model corresponding to the business type to which the data subset belongs includes:

Train multiple user perception profiling models based on data selected from this data subset;

The accuracy of each trained user perception profile model is counted, and the user perception profile model with the highest accuracy is selected as the user perception profile model corresponding to the business type to which the data subset belongs.

Specifically, the step of obtaining the user's network awareness description corresponding to the network awareness score also includes:

If it is learned that the target user has a poor network perception problem according to the user network perception description corresponding to the network awareness score, then obtain the duration of the poor network perception problem according to the timestamp in the perception fusion information of the target user;

The location information of the target user within the duration and the occupied network element information or cell information are obtained from the perceptual fusion information of the target user.

Specifically, after the step of obtaining the location information of the target user within the duration and the occupied network element information or cell information from the perceptual fusion information of the target user, the step further includes:

Determine the severity of the poor network perception problem according to the importance of the service type of the target user's perception fusion information, user star rating and the duration of the poor network perception problem;

Determine the importance level of the network element or cell occupied by the target user based on the number of users, traffic volume, and scenario labels of the network element occupied by the target user, or the number of users, traffic volume, and scenario labels of the cell;

Weight the severity of the poor network perception problem and the importance level of the network element or cell occupied by the target user to obtain the processing priority of the poor network perception problem;

If the processing priority reaches a preset threshold, the network element or cell occupied by the target user is investigated to determine the cause of the poor network perception problem.

According to a second aspect of an embodiment of the present invention, a user network perception profiling device is provided, which includes:

A collection module, used to collect the target user's online behavior data, XDR data, MR data and complaint data, and parse the performance data that characterizes the target user's network perception from the online behavior data;

An association module, used to associate the performance data, XDR data, MR data and complaint data, obtain the perceptual fusion information of the target user, and obtain the service type of the perceptual fusion information according to the perceptual fusion information;

A profiling module, configured to select data corresponding to the service type from the perceptual fusion information according to the service type; input the selected data into the user perceptual profiling model, output the network perceptual score of the target user, and obtain the User network perception description corresponding to network perception score;

Wherein, the service type and the data corresponding to the service type are stored in association in advance;

The network awareness score and the user network awareness description are stored in association in advance;

The user perception portrait model is trained and acquired based on the perception fusion information of the user sample and the network perception score of the user sample.

According to a third aspect of the embodiment of the present invention, an electronic device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor can execute the program instructions by calling the program instructions. The user network perception profiling method provided by any of the various possible implementation methods of the first aspect.

According to a fourth aspect of the embodiment of the present invention, a non-transitory computer-readable storage medium is also provided, the non-transitory computer-readable storage medium stores computer instructions, the computer instructions cause the computer to execute the first aspect The user network perception profiling method provided by any of the various possible implementation methods.

Embodiments of the present invention provide a user network perception profiling method and device. The method collects user online behavior data, XDR data, MR data and complaint data, and parses the user online behavior data to obtain performance data that characterizes the user's network perception. Correlate with XDR data, MR data and complaint data, and select corresponding data from the associated data according to the service type of user network perception for network perception scoring, thereby achieving an accurate user network perception portrait.

Description of the drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the drawings in the following description These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without exerting creative efforts.

Figure 1 is a schematic flow chart of a user network awareness profiling method provided by an embodiment of the present invention;

Figure 2 is a schematic flow chart of the implementation principle of the user network sensing profiling method provided by an embodiment of the present invention;

Figure 3 is a schematic diagram of the data collection process in the user network perception profiling method provided by an embodiment of the present invention;

Figure 4 is a schematic diagram of the data processing flow in the user network perception profiling method provided by an embodiment of the present invention;

Figure 5 is a schematic flowchart of performing user perception profiling based on fused data in the user network perception profiling method provided by an embodiment of the present invention;

Figure 6 is a schematic diagram of the process of locating the problem of poor user network perception in the user network perception profiling method provided by an embodiment of the present invention;

Figure 7 is a schematic structural diagram of a user network sensing profiling device provided by an embodiment of the present invention;

FIG. 8 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

Detailed ways

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the drawings in the following description These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without exerting creative efforts.

In one embodiment of the present invention, a user network-aware profiling method is provided. Figure 1 is a schematic flow chart of a user network-aware profiling method provided by an embodiment of the present invention. The method includes: S101, collecting the target user's online behavior data and XDR data. , MR data and complaint data, and parse the performance data that characterizes the network perception of the target user from the online behavior data;

As shown in Figure 2, the implementation principle of this embodiment is mainly to automatically collect user online behavior data, associate and match it with XDR data, complaint data, MR fingerprint database, etc., and establish a second-level granular user perception fusion data information database. Through big data analysis and modeling, user perception profiling is carried out, and further automatic and precise positioning of poor perception quality problems and user perception preventive maintenance are realized.

With the development of communication technology, electronic devices occupy a large amount of users' time. While users receive a large amount of information provided by the network, they also leave behind massive trace data. These traces can effectively reflect the user's true perception. In order to accurately assess user network perception and achieve precise location of user network perception portraits and poor perception quality issues, effective data collection is crucial, which directly affects the accuracy and completeness of subsequent analysis. This embodiment needs to collect the target user's online behavior data, user XDR (External Data Representation, external data representation) data, complaint data and MR (Measurement Report, measurement report) data. Among them, the target users are users who need network awareness profiling.

The flow chart of data collection is shown in Figure 3. Among them, user online behavior data includes various behavioral data generated by users using network functions through APP (Application, application) and web pages, and the granularity can reach the second level. More specifically, the user's online behavior data collection path mainly includes log data collection and network data collection. Among them, log data collection can be automatically collected using methods such as SDK buried points; network data collection mainly uses web crawlers or public APIs (Application Programming Interface, application program interfaces) to capture the required network data. The collected user online behavior data mainly includes: username, page view, unique visitor ID, receive data traffic, send data traffic transmit, packet ID PID, log, average usage time, and total usage time. , network type and operator, failure to send messages, number of visits, number of people, duration and error rate of functions and interfaces, number of messages sent and received, number of phone calls initiated and received, number of VoicemailPlay/Delete times, login duration, number of failed logins, Fields such as the number of message sending failures and phone initiation and reception failures, phone call quality, IMEI (International Mobile Equipment Identity, International Mobile Equipment Identity), device model, longitude, latitude, website URL, submission information post, etc.

User XDR data is realized through platform docking. The collected XDR data mainly includes S1-MME and S1-U/S11 original code streams, that is, the user's Local Province, the user's city Local City, roaming type, interface, xDR ID, IMEI, Procedure Start Time, Procedure End Time, Procedure Status, Cell ID, MME UE S1AP ID, IMSI (International Mobile Subscriber Identity, International Mobile Subscriber Identity), longitude, latitude, App Type, App Sub-type, App Content, App Status, Operation Type, Login-Success, Login-Request-Time, Login-Response-Time, Uplink Data UL Data, Downlink Data DL Data , URI((Uniform ResourceIdentifier, Uniform Resource Identifier), HTTP response time HTTP (HyperText Transfer Protocol, Hypertext Transfer Protocol) Response Time, host name HOST, HTTP content type HTTP_content_type, service component flag Service Comp Flag, service behavior flag Service Behavior Flag, service time Service Time, video download time VideoDownTime, initial buffer duration InitbufferDuration, uplink duration updura, downlink duration downdura, UL disorder IP packet UL Disorder IP Packet, DL disorder IP packet DLDisorder IP Packet, UL re Transmission IP packet UL Retrans IP Packet, DL retransmission IP packet DLRetrans IP Packet, TCP response time TCP (Transmission Control Protocol, Transmission Control Protocol) Response Time, TCP confirmation time TCP ACK Time, request time Req Time, UL_AVG_RTT, DW_AVG_RTT and other fields .

MR data is realized through platform docking, including user-level and community-level MR. The user-level MR collection data mainly includes timestamp Time, MME UE S1AP ID, location-longitude, location-latitude, eNB ID, Cell ID, MR type, MR type, AoA (Angle-of-Arrival, angle of arrival) Ranging), Serving RSRP (Reference Signal Received Power, Reference Signal Received Power), CQI (ChannelQuality Indication, Channel Quality Indication), neighboring cell information, etc. Cell-level MR collection data mainly includes eNB ID, Cell ID, timestamp Time, eNB Received Interfere, UL Packet Loss, DL Packet Loss, air interface user plane uplink service bytes, and air interface user plane downlink Fields such as the number of service bytes, PRB occupancy rate of uplink and downlink service channels, and the number of simultaneous online users.

User complaint data collection is realized through platform docking. The collected data mainly includes fields such as complaint type, complaint time, IMEI, and device type. The data processing unit mainly performs operations such as analysis, format conversion, and normalization processing of the massive data collected in the above steps. Store the data obtained in the above steps that can be used to analyze the target user's network perception in a data storage unit.

The existing platform conditions for the data sources that need to be collected can be met. After the platform docking is completed, no additional data collection related work is required. The existing platform automatically pushes the required collected data, saving a lot of manpower and equipment costs.

Parse the performance data that characterizes the target user's network perception from the collected user online behavior data, mainly including response time data, frame rate and fluency data, such as SM (application drawing polling frequency), FPS (Frames Per Second, per second) The frame rate of the clock-filled image) and SF (Skipped Frames, the number and amplitude of application skipped frames), as well as log data, etc.

S102, associate the performance data, XDR data, MR data and complaint data to obtain the perceptual fusion information of the target user, and obtain the service type of the perceptual fusion information according to the perceptual fusion information;

As shown in Figure 4, before subsequent analysis, it is necessary to clean and fuse the collected user sample data to build a user perception fusion database. Use the user perception fusion database to train the user perception portrait model to establish an accurate user perception portrait model. Correlate the collected data of the target user to obtain the target user's perception fusion information, and use the perception fusion information to profile the target user's network perception based on the trained perception portrait model.

The purpose of data cleaning is to eliminate the impact of outliers, missing values, etc. on subsequent analysis. Optionally, methods such as direct deletion and mean correction can be used to deal with outliers, and regression interpolation, Lagrangian interpolation, multiple imputation methods, etc. can be used to impute missing values.

S103: Select data corresponding to the service type from the perception fusion information according to the service type; input the selected data into the user perception portrait model, output the network perception score of the target user, and obtain the network perception score Corresponding user network awareness description;

The user perception portrait model can be a support vector machine, a neural network, a random forest, etc. This embodiment is not limited to the type of user perception portrait model.

Wherein, the service type and the data corresponding to the service type are stored in association in advance; the network perception score and the user network perception description are stored in association in advance; the user perception portrait model is based on the perception fusion information of user samples and the The network perception scores of user samples are obtained through training.

Common phenomena that affect users' network perception mainly include freezes and crashes. This embodiment can parse second-level service fluency through user online behavior data, such as SM data, and further combine with user complaint data to perform perceived satisfaction ratings based on user online behavior data, and store the perceived satisfaction rating data of user samples in User perception fusion information base. Due to the different characteristics of various services, different user behaviors have obviously different requirements for network quality. Therefore, this embodiment establishes a user-perceived satisfaction rating standard based on different user behaviors, as shown in Table 1.

Table 1 User’s network perception scoring criteria

This embodiment collects user online behavior data, XDR data, MR data and complaint data, and parses the user online behavior data to correlate the performance data representing the user's network perception with the XDR data, MR data and complaint data. According to the user Network-aware business types select corresponding data from associated data for network-aware scoring, thereby achieving an accurate user network-aware portrait.

On the basis of the above embodiment, in this embodiment, the performance data, XDR data, MR data and complaint data are associated, and the step of obtaining the perceptual fusion information of the target user includes: according to the XDR data, MR data and the timestamp, IMEI, IMSI, MME UE S1AP ID, Cell ID and eNB ID fields that all exist in the complaint data, associate the XDR data, MR data and complaint data to obtain the associated fusion data of the target user;

The data of target users and user samples are all correlated, and finally the correlation and fusion data including timestamps, GPS (Global Positioning System, Global Positioning System) level longitude and latitude information, occupied cell information, network performance indicators, user information and user behavior information are obtained. Among them, user information includes IMEI, IMSI, device model and IP, etc., and user behavior information includes URL, log, PID, etc.

According to the timestamp, App Type, App Sub-type and IMEI fields that exist in the performance data and the associated fusion data, the performance data and the associated fusion data are associated to obtain the perceptual fusion information of the target user.

The finally obtained perceptual fusion information of the target user and user sample is user-level network perceptual fusion information with a granularity of up to seconds.

On the basis of the above embodiment, in this embodiment, the selected data is input into the user perception portrait model, and the step of outputting the network perception score of the target user also includes: obtaining each user according to the perception fusion information of each user sample. The service type of the perceptual fusion information of the sample; according to the service type of the perceptual fusion information of all the user samples, the perceptual fusion information of all the user samples is divided into multiple data subsets; for any of the data subsets, According to the business type to which the data subset belongs, data corresponding to the business type is selected from the data subset; the user perception portrait model is trained according to the data selected from the data subset to obtain the data to which the data subset belongs. User perception portrait model corresponding to business type;

This embodiment establishes a service-level and user-level perceptual correlation fusion information database. As shown in Figure 5, user perception portrait model training for different services is performed based on the user perception fusion information database. The service type of perception fusion information can be distinguished by the AppType, App Sub-type and PID fields. Specifically, the user perception score is used as a prediction variable, that is, the output of the user perception profile model, and different types of business data are divided into multiple data subsets. For each of the subsets, network performance indicators, device types, and user XDR data suitable for the service type can be selected as training data for training the user perception profile model.

Correspondingly, the step of inputting the selected data into the user perception profile model and outputting the network perception score of the target user includes: according to the service type of the target user's perception fusion information, obtaining the user perception profile model corresponding to the service type. ; Input the data selected from the target user's perception fusion information into the user perception portrait model corresponding to the service type, and output the network perception score of the target user.

When performing network perception scoring based on the perceptual fusion information of the target user, the corresponding user perceptual portrait model is selected for analysis based on the type of perceptual fusion information.

This embodiment establishes a second-level perceptual correlation fusion information database at the business and user levels. Compared with traditional solutions that rely on network element measurement indicators, this embodiment is more accurate and more suitable for user experience. The perceptual correlation fusion information database contains massive user behavior data. The samples are rich and can avoid the problem of insufficient representation mechanism of network element-level indicators; an accurate user perception portrait model is constructed through big data and machine learning, thereby improving the accuracy of user network perception portraits.

On the basis of the above embodiment, in this embodiment, the user perception profile model is trained based on the data selected from the data subset. The step of obtaining the user perception profile model corresponding to the business type to which the data subset belongs includes: Train multiple user perception portrait models based on the data selected from the data subset; count the accuracy of each trained user perception profile model, and select the user perception profile model with the highest accuracy as the data subset to which it belongs The user perception portrait model corresponding to the business type.

Specifically, each data subset is used to simultaneously train multiple user perception portrait models, such as support vector machines, neural networks, and random forests. Calculate the accuracy of each model, etc., and select the user perception profile model with the best comprehensive performance as the model for user network perception profile.

On the basis of the above embodiments, in this embodiment, the step of obtaining the user network awareness description corresponding to the network awareness score further includes: if it is learned that the target user exists according to the user network awareness description corresponding to the network awareness score If the network perception difference problem is detected, the duration of the network perception difference problem is obtained according to the timestamp in the perception fusion information of the target user; and the target user within the duration is obtained from the perception fusion information of the target user. Location information, such as latitude and longitude information, and occupied network element information or cell information.

Specifically, after obtaining the user perception portrait model, user-level accurate perception evaluation with second granularity can be achieved. For users with poor perception, it is necessary to further locate the quality problem that leads to poor perception for subsequent processing. The implementation steps are shown in Figure 6. The obtained location information, as well as occupied network element information or cell information are output as location analysis data for poor perception problems.

On the basis of the above embodiment, as shown in Figure 6, in this embodiment, the location information of the target user within the duration and the occupied network element information or cells are obtained from the perceptual fusion information of the target user. The step of information further includes: judging the severity of the poor network perception problem based on the importance of the service type of the target user's perception fusion information, the user's star rating and the duration of the poor network perception problem;

Determine the importance level of the network element or cell occupied by the target user based on the number of users, traffic volume, and scenario labels of the network element occupied by the target user, or the number of users, traffic volume, and scenario labels of the cell;

Weight the severity of the poor network perception problem and the importance level of the network element or cell occupied by the target user to obtain the processing priority of the poor network perception problem;

If the processing priority reaches the preset threshold, the network element or cell occupied by the target user is checked by checking key indicators, work parameters and other means to determine the cause of the poor network perception problem. At the same time, based on GPS-level precise location information, the geographical presentation of poor quality communities is realized.

Through the trained user perception portrait model, user perception assessment portraits for different services can be accurately realized in seconds. Further combined with the GPS-level precise location information in the fusion information database, it can dynamically reflect the perception situation of the entire network and GIS (Geographic Information System, geography). Information system) presentation. In addition, the causes of poor user perception and related network elements or communities can be further located in a timely manner, maintenance processing information can be generated and pushed to relevant units and departments, and user care information or processing progress information can also be sent in a timely manner through various network media.

Since the user perception fusion database in this embodiment can provide GPS-level location information, location positioning can be achieved more accurately. Compared with the traditional solution that only relies on user complaints and work parameters, the accuracy is higher.

Another embodiment of the present invention provides a user network perception profiling device, which is used to implement the methods in the foregoing embodiments. Therefore, the descriptions and definitions in the foregoing embodiments of the user network awareness profiling method can be used to understand each execution module in the embodiments of the present invention. Figure 7 is a schematic diagram of the overall structure of a user network perception profiling device provided by an embodiment of the present invention. The device includes a collection module 701, an association module 702 and a profiling module 703; wherein,

The collection module 701 is used to collect the target user's online behavior data, XDR data, MR data and complaint data, and parse the performance data that characterizes the target user's network perception from the online behavior data;

The association module 702 is used to associate the performance data, XDR data, MR data and complaint data, obtain the perceptual fusion information of the target user, and obtain the service type of the perceptual fusion information according to the perceptual fusion information;

The profiling module 703 is configured to select data corresponding to the service type from the perception fusion information according to the service type; input the selected data into the user perception profile model, output the network perception score of the target user, and obtain the User network perception description corresponding to network perception score;

Wherein, the service type and the data corresponding to the service type are stored in association in advance; the network perception score and the user network perception description are stored in association in advance; the user perception portrait model is based on the perception fusion information of user samples and the The network perception scores of user samples are obtained through training.

This embodiment collects user online behavior data, XDR data, MR data and complaint data, and parses the user online behavior data to correlate the performance data representing the user's network perception with the XDR data, MR data and complaint data. According to the user Network-aware business types select corresponding data from associated data for network-aware scoring, thereby achieving an accurate user network-aware portrait.

Figure 8 illustrates a schematic diagram of the physical structure of an electronic device. As shown in Figure 8, the electronic device may include: a processor (processor) 801, a communications interface (Communications Interface) 802, a memory (memory) 803, and a communication bus 804. Among them, the processor 801, the communication interface 802, and the memory 803 complete communication with each other through the communication bus 804. The processor 801 can call the logical instructions in the memory 803 to perform the following method: collect the target user's online behavior data, XDR data, MR data and complaint data, and parse out the characteristics of the target user from the online behavior data. Network-perceived performance data; associate the performance data, XDR data, MR data and complaint data to obtain the perceptual fusion information of the target user, and obtain the service type of the perceptual fusion information according to the perceptual fusion information; Select the data corresponding to the service type from the perception fusion information according to the service type; input the selected data into the user perception portrait model, output the network perception score of the target user, and obtain the network perception score corresponding to the network perception score. User network awareness description.

In addition, the above-mentioned logical instructions in the memory 803 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention essentially or the part that contributes to the existing technology or the part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code. .

This embodiment provides a non-transitory computer-readable storage medium. The non-transitory computer-readable storage medium stores computer instructions. The computer instructions cause the computer to execute the methods provided by the above method embodiments. For example, it includes: collecting the Internet access of the target user. Behavior data, XDR data, MR data and complaint data, and parse the performance data representing the network perception of the target user from the Internet surfing behavior data; correlate the performance data, XDR data, MR data and complaint data , obtain the perceptual fusion information of the target user, and obtain the service type of the perceptual fusion information according to the perceptual fusion information; select the data corresponding to the service type from the perceptual fusion information according to the service type; The selected data is input into the user perception portrait model, the network perception score of the target user is output, and the user network perception description corresponding to the network perception score is obtained.

Those of ordinary skill in the art can understand that all or part of the steps to implement the above method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, It includes the steps of the above method embodiment; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other various media that can store program codes.

The device embodiments described above are only illustrative. The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in One location, or it can be distributed across multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Persons of ordinary skill in the art can understand and implement the method without any creative effort.

Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the part of the above technical solution that essentially contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., including a number of instructions to cause a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments or certain parts of the embodiments.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it can still be used Modifications are made to the technical solutions described in the foregoing embodiments, or equivalent substitutions are made to some of the technical features; however, these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (8)

1. A user network perception profiling method, which is characterized by including: Collect the target user's online behavior data, XDR data, MR data and complaint data, and parse the performance data that characterizes the target user's network perception from the online behavior data; Correlate the performance data, XDR data, MR data and complaint data to obtain the perceptual fusion information of the target user, and obtain the service type of the perceptual fusion information according to the perceptual fusion information; Select the data corresponding to the service type from the perception fusion information according to the service type; input the selected data into the user perception portrait model, output the network perception score of the target user, and obtain the network perception score corresponding to the network perception score. User network awareness description; Wherein, the service type and the data corresponding to the service type are stored in association in advance; The network awareness score and the user network awareness description are stored in association in advance; The user perception portrait model is trained and acquired based on the perception fusion information of the user sample and the network perception score of the user sample; The step of obtaining the user's network awareness description corresponding to the network awareness score also includes: If it is learned that the target user has a poor network perception problem according to the user network perception description corresponding to the network awareness score, then obtain the duration of the poor network perception problem according to the timestamp in the perception fusion information of the target user; Obtain the location information of the target user within the duration and occupied network element information or cell information from the perceptual fusion information of the target user; The step of obtaining the location information of the target user within the duration and the occupied network element information or cell information from the perceptual fusion information of the target user also includes: Determine the severity of the poor network perception problem according to the importance of the service type of the target user's perception fusion information, user star rating and the duration of the poor network perception problem; Determine the importance level of the network element or cell occupied by the target user based on the number of users, traffic volume, and scenario labels of the network element occupied by the target user, or the number of users, traffic volume, and scenario labels of the cell; Weight the severity of the poor network perception problem and the importance level of the network element or cell occupied by the target user to obtain the processing priority of the poor network perception problem; If the processing priority reaches a preset threshold, the network element or cell occupied by the target user is investigated to determine the cause of the poor network perception problem. 2. The user network perception profiling method according to claim 1, characterized in that the step of collecting the target user's online behavior data includes: Collect log data of the target users through SDK burying method; Crawl the network data of the target user through a web crawler or public API; The collected log data and captured network data are used as the target user's online behavior data. 3. The user network perception profiling method according to claim 1, characterized in that the step of correlating the performance data, XDR data, MR data and complaint data to obtain the perception fusion information of the target user includes: According to the timestamp, IMEI, IMSI, MME UE S1APID, CellID and eNBID fields that exist in the XDR data, MR data and complaint data, associate the XDR data, MR data and complaint data to obtain the target user associated fusion data; According to the timestamp, App Type, App Sub-type and IMEI fields that exist in the performance data and the associated fusion data, the performance data and the associated fusion data are associated to obtain the perceptual fusion information of the target user. 4. The user network awareness profiling method according to claim 1, wherein the step of inputting the selected data into the user awareness profiling model and outputting the network awareness score of the target user further includes: The business type of obtaining the perceptual fusion information of each user sample based on the perceptual fusion information of each user sample; According to the service type of the perceptual fusion information of all the user samples, divide the perceptual fusion information of all the user samples into multiple data subsets; For any of the data subsets, select the data corresponding to the business type from the data subset according to the business type to which the data subset belongs; Train the user perception profile model based on the data selected from the data subset, and obtain the user perception profile model corresponding to the business type to which the data subset belongs; Accordingly, the step of inputting the selected data into the user perception portrait model and outputting the network perception score of the target user includes: According to the service type of the target user's perception fusion information, obtain the user perception portrait model corresponding to the service type; The data selected from the perceptual fusion information of the target user is input into the user perceptual portrait model corresponding to the service type, and the network perceptual score of the target user is output. 5. The user network-aware profiling method according to claim 4, characterized in that, the user-aware profiling model is trained according to the data selected from the data subset, and the users corresponding to the business types to which the data subset belongs are obtained. The steps of perceptual portrait model include: Train multiple user perception profiling models based on data selected from this data subset; The accuracy of each trained user perception profile model is counted, and the user perception profile model with the highest accuracy is selected as the user perception profile model corresponding to the business type to which the data subset belongs. 6. A user network-aware profiling device, characterized in that executing the user network-aware profiling method according to any one of claims 1 to 5 includes: A collection module, used to collect the target user's online behavior data, XDR data, MR data and complaint data, and parse the performance data that characterizes the target user's network perception from the online behavior data; An association module, used to associate the performance data, XDR data, MR data and complaint data, obtain the perceptual fusion information of the target user, and obtain the service type of the perceptual fusion information according to the perceptual fusion information; A profiling module, configured to select data corresponding to the service type from the perceptual fusion information according to the service type; input the selected data into the user perceptual profiling model, output the network perceptual score of the target user, and obtain the User network perception description corresponding to network perception score; Wherein, the service type and the data corresponding to the service type are stored in association in advance; The network awareness score and the user network awareness description are stored in association in advance; The user perception portrait model is trained and acquired based on the perception fusion information of the user sample and the network perception score of the user sample. 7. An electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that when the processor executes the program, it implements any one of claims 1 to 5 The steps of the user network awareness profiling method described in the item. 8. A non-transitory computer-readable storage medium with a computer program stored thereon, characterized in that when the computer program is executed by a processor, the user network-aware profiling method according to any one of claims 1 to 5 is implemented. step.